J. Physiol. (1997)  505P: 78

Visual stabilisation during dynamic head roll in avestibular human subjects.

A.R. Gardner-Medwin, M. Doshi, *M. Faldon and *A. Bronstein
Dept. Physiology, UCL, London WC1E 6BT and *MRC HMBU, Queen Square, London WC1N 3BG

We have measured the effects of voluntary oscillating head roll (about an occipito-nasal axis) on ocular compensation and on the apparent movement of the visual world in three chronically avestibular subjects.

Movement of the eyes in the head was measured by video-oculography (Scherer et al., 1991) with head roll amplitudes of 25-40° (peak to peak). Ocular counter-roll was a small fraction of the head movements, 3% to 8% at 0.33 Hz and -2% to 8% at 1Hz, with few torsional saccades. This compares with a gain of 40-70% between saccades in normals (Collewijn et al., 1985).

The patients were little impaired in setting a static subjective visual vertical (SVV) in the dark. Only one of the three set a consistent tilt (3.7° ± 0.2° S.E.M., N=30) outside the 95% confidence limits (± 2.6°) based on 8 normal subjects. The average S.D. for settings from random starting angles was 0.99° in the patients and 0.74° in normals. The effects of static 20-40° head tilt on SVV (Müller effect) were similar, with the set line having an actual tilt of -12% ± 11% of head tilt in the patients (M ± S.E.M., N=3) and -11% ± 4% in normals (N=8). In neither group was SVV significantly affected by a 10N sideways force applied to a band around the head.

The effect of head roll on stability of the visual world was assessed with apparatus that measured the angle of head tilt and varied the tilt of a stimulus (a line of dots, viewed otherwise in the dark) in proportion to this angle (Gardner-Medwin et al., 1996). The subject adjusted the gain of this relationship until the line appeared stable. For the patients this gain was 24% ± 10% at 0.33Hz and 32% ± 15% at 1Hz (M ± S.E.M., N=3) compared with normals: 8.9% ± 2.1% at 0.33Hz and 7.0% ± 2.9% at 1Hz (N=8). Note that these gains are positive (opposite to the Müller effect). With oscillating roll, unlike static displacement, subjects generally (excepting one normal subject) required the stimulus to move in the same direction as head tilt to appear stable. This effect was larger in avestibular subjects, but less than the 92-102% gain that would stabilise the the retinal image based on the measured counter-roll. The patients were less consistent in their adjustments: the mean S.D. for repeated settings was 20% for patients and 9% for normals. We conclude that perceptual instabilities during dynamic head roll are more sensitive to vestibular loss than static measurements, but that avestibular subjects do compensate in other ways.

Supported by the MRC. The study has ethical approval.


Collewijn H., Van der Steen J., Ferman L., & Jansen T.C. (1985) Exp. Brain. Res. 59:185-196

Gardner-Medwin, A.R., Varughese M. and Margetts, P. (1996) J.Physiol. 494:2-3P

Scherer H, Teiwes W & Clarke AH (1991) Acta Otolaryngol. 111:182-187